Platelet Activating Factor and Convolutions of the Brain

When I was a child, there was a popular detective show on TV in which each detective had a catchy nickname. On the basis of these nicknames, we were able to form an image in our minds of each character's personality. A problem occurred, however, when the image of each of the characters became fixed, thus making it difficult for the writers to come up with new story lines. The only way the show could develop was to kill off its detectives so that new characters could take their place. When an object is given a name that conjures up a certain mental image, people unfamiliar with it may develop fixed and inflexible concepts about it.

In the late 1970s, a substance that strongly activates platelets was isolated from basophils. This substance, which was named "platelet activating factor" (PAF) was neither a protein nor a water-soluble, low molecular weight compound, but it was a substance with a phospholipid backbone. Based on the fact that PAF strongly activates platelets and neutrophils, it was easy to assume that it probably plays a major role in allergies and anaphylactic shock. Therefore, diligent research was conducted in those fields on PAF and its metabolizing enzymes. Recently, clinical applications have also been investigated for a substance called PAF antagonist and a PAF-specific hydrolytic enzyme (PAF acetylhydrolase, PAF-AH).

Early in the 1980s, it was reported that enzymatic activity concerning the synthesis and degradation of PAF is present in the brain and that neurons exhibit a variety of responses to PAF. However, almost no in-depth studies have been performed since then concerning the physiological significance of this finding. Various reasons for this can be proposed, such as the fact that PAF is difficult to handle, detect, and quantify because it is a phospholipid, and that special techniques are required. I believed, however, that the main reason lies in the name of this mediator: platelet activating factor. This name attracted the interest of hematologists and immunologists, but most researchers in other fields probably thought: Platelet activating factor? Phospholipid? This substance is not related to my research. I believe, however, that it is unnatural to think PAF and PAF-metabolizing enzymes are present in the brain, but surely they are not related to brain function.

When I began research as a graduate student, I selected "purification and characterization of the PAF-specific hydrolytic enzyme (PAF-AH) present in the brain" as my topic. I did so for two reasons. The first was that diligent research on mammalian phospholipase A2 was being conducted in my laboratory at that time, so I was naturally interested in this general field. The second reason was that more than 10 years had passed since this enzyme had been discovered in the brain, but no one was conducting serious research on it. Although it had been reported that PAF and PAF receptors are present in the brain, there had been almost no inquiry into the significance of this finding. Therefore, I purified PAF-AH and cloned its gene because I believed that by revealing its structure and activity regulating mechanism, we could obtain clues to understanding the function of PAF in the brain.

First, I discovered that there are several PAF-AH isoforms present in mammalian brain cells. I then successfully purified the enzyme with the most important activity (1). I also discovered that this enzyme has a heteromeric structure comprising three subunits and showed that one of these subunits is not necessary for enzymatic activity (1, 2, 3). Next, I performed cDNA cloning of the three subunits and obtained some very interesting results. First, the amino acid sequences of the two catalytic subunits were very similar, but because they have absolutely no homology with other known proteins, they do not belong to any previously know enzyme subgroup, and that suggested they have a novel three-dimensional structure (4, 5).

Moreover, I discovered that the subunit not needed for activity (which I have tentatively called the regulatory subunit) is, surprisingly, the product of the gene that has been linked to Miller-Dieker syndrome, a congenital morphogenetic hypoplasia of the brain in humans (6). In Miller-Dieker syndrome, neuron migration during brain development is incomplete, and the brains of patients with this condition have smooth surfaces with no gyra (commonly called convolutions). My findings demonstrate that brain convolutions are not formed if the PAF-AH regulatory subunit is missing. There has been little progress in research on intracellular signaling systems during neuron migration, and it is still unclear how PAF-AH (and probably PAF as well) is involved. I have discovered, however, that the catalytic subunits and the regulatory subunit of this enzyme are dissociated in vitro by heparin (6), that this dissociation is reversible, and that both unbound catalytic subunits and catalytic subunits bound to the regulatory subunits are present in brain cells (7). These findings suggest that in brain cells the association and dissociation of the regulatory and catalytic subunits is controlled by some unknown factor. Moreover, I believe that in all probability the concentration of PAF present in the cells is determined by this factor and that this leads to a signal further downstream. In collaboration with other researchers, I have recently performed an X-ray crystallographic analysis of the catalytic subunits. Our findings have revealed that the catalytic subunit has a three-dimensional structure completely different from lipase and esterase groups, and instead has a structure similar to the alpha-subunit of the trimeric G-proteins (8). On the other hand, a characteristic repeating structure called a WD repeat is present on the regulatory subunit, and this structure is also present on the beta-subunit of trimeric G proteins. When we combine this information, we can conclude that the intracellular PAF-AH has a trimeric G-protein type structure (8). This structure suggests that the enzyme undergoes dynamic subunit assembly and disassembly in the cell, and that this activity is important in the regulation of some kind of function.

I have been involved in the detection of activity, purification, antibody preparation, cDNA cloning and X-ray crystallography of the PAF-AH enzyme. Moreover, knockout mice for each of the enzyme's subunits are expected to be born later this year. Of course, it is still unclear exactly how PAF and PAF-AH are involved in brain morphogenesis. However, there is only a short history of research on intracellular signaling in the developing nervous system and at present there is no reason to view our lack of understanding in this area with pessimism. I am certain that in the near future we will find the name of PAF-AH on a diagram depicting the new signaling systems in the brain. Before that, however, the name of "platelet activating factor" may have to be

References

1. Hattori, M., Arai, H., Inoue, K., J. Biol. Chem. 268, 18748 (1993).
2. Hattori, M., Adachi, H., Tsujimoto, M., Arai, H., Inoue, K., Nature 370, 216 (1994).
3. Inoue, K., Arai, M., Hattori, M., J. Lipid Mediators Cell Signaling 10, 13 (1994).
4. Hattori, M., Adachi, H., Tsujimoto, M., Arai, H., Inoue, K., J. Biol. Chem. 269, 23150 (1994).
5. Hattori, M. et al., J. Biol. Chem. 270, 31345 (1995).
6. Hattori, M., Adachi, H., Tsujimoto, M., Arai, H., Inoue, K., Nature 370, 216 (1994).
7. M. Hattori, unpublished material.
8. Ho, Y.S. et al., Nature 385, 89 (1997).

To Advertise   |   Find Products